Dr. Hinchey proposes to develop the skills necessary to become an independent researcher in the field of quality assessment for cerebrovascular disease. The candidate has shown a commitment to clinical research by both her training as a cerebrovascular fellow at the Cleveland Clinic Foundation and her experience as co-director of an American Academy of Neurology sponsored Acute Stroke Outcomes Pilot Project. The candidate outlines the courses she will take over the first four years of the grant to develop skills in research design, methodology, statistics, data management, health policy, cost effectiveness and ethics in research. Her long-term aim is to develop methods to assess the quali1y of care for stroke patients in both the inpatient and oupatient setting. An integral component of this is the ability to compare patients from different settings. The proposed project, to be done over the five years of the grant, develops ways to compare patients. The project's specific aim is to develop risk-adjusted models that predict both in-hospital morbidi1y and mortality for stroke patients based on admission characteristics. The proposed project will address four short-comings noted in current risk-adjusted stroke mortality prediction models; 1) the use of mortality alone as a measure of quality, 2) the paucity of neurologic specific risk variables in current models, 3) the inclusion of hemorrhagic and ischemic stroke into one model, and 4) the lack of process of care measurements in current models. Potential predictors of outcomes will be prospectively collected on all stroke patients hospitalized at Strong Memorial Hospital and seen by the neurology department. Outcome data will be collected at discharge, 1 month and 3 months. The models will be developed based on 2-Y2 years of patient data collection. The model will then be validated on another group of patients collected over 1 -V2 years. The final models developed will be used as the preliminary data for future research to extensively validate the models ability to measure quality of care. Then, the comparison of predicted versus actual outcomes can be used by hospitals or payers to assess quality. Supervision for this project will be by a multi disciplinary team at the University of Rochester, University of California and the Cleveland Clinic Foundation with experience in fostering the development of clinical researches.